economic shock, owner-manager incentives, and corporate … version... · 2010-09-18 ·...
TRANSCRIPT
Economic Shock, Owner-Manager Incentives, and Corporate Restructuring: Evidence from the Financial Crisis in Korea1
Jun-Koo Kanga,*, Inmoo Leeb, and Hyun Seung Nac
a Division of Banking and Finance, Nanyang Business School, Nanyang Technological University, Singapore b Dimensional Fund Advisors, Austin, Texas 78746, USA c Department of Economics and Finance, College of Business, City University of Hong Kong, Hong Kong
November 2009
JEL Classification: G30, G32, G34
Keywords: Owner-manager incentives, Corporate restructuring, Financial crisis, Cash flow and control rights.
* Corresponding author: Tel.: +65 6790 5662; fax: +65 6791 3697 E-mail address: [email protected] (J.-K. Kang), [email protected] (I. Lee), [email protected] (H. S. Na).
Economic Shock, Owner-Manager Incentives, and Corporate Restructuring:
Evidence from the Financial Crisis in Korea
Abstract
We examine how owner-managers incentives and firm-specific measures of corporate governance affect restructuring decisions during an economy-wide shock. Using a large sample of Korean firms that had experienced a severe financial crisis during 1997-1998, we find that the likelihood of restructuring is negatively related to the divergence of cash flow rights and control rights of controlling shareholders, and that the announcements of restructuring by chaebol firms with such divergence are greeted more negatively by investors. However, firm-specific measures of corporate governance such as total debt, bank loans, and equity ownership by unaffiliated financial institutions mitigate these negative effects, thereby influencing firms to choose value-maximizing restructuring policies. Our results suggest that the controlling shareholders’ incentives to expropriate other investors are high during an economic shock. Our results also highlight the importance of corporate governance in mitigating such expropriation incentives, and provide important implications for the role of corporate governance during an economic shock, such as the 2007-2008 global financial crisis.
1
1. Introduction
Market-wide shocks in the economy, such as the world-wide financial crisis in 2007-2008 caused by
the U.S. subprime mortgage problems, the Great Depression in the U.S., and the financial crisis of East
Asia in 1997, have a profound negative impact on asset prices, firms’ investment and financing policies,
investor sentiments, and consumer demands. To overcome these economy-wide adverse effects, firms
that experience a market-wide shock (hereafter referred to as an economic shock) are expected to
undertake several restructuring measures, which might be fundamentally different from those undertaken
by firms that experience a firm-specific shock. However, evidence on the potential difference in
restructuring measures during these different economic environments is limited. In particular, although
there is extensive literature on the restructuring activity responding to firm-specific shocks, such as
performance declines,1
This lack of empirical evidence on the restructuring activity in response to an economic shock is
unfortunate, given that the ability of firms to respond to an unexpected economic shock and the
determinants of this response are important factors that affect firms’ organizational efficiency.
Furthermore, an economic shock tends to create different incentives among controlling shareholders,
outside shareholders, and creditors concerning restructuring policies, since the agency problems among
investors tend to increase significantly during the shock. For example, Johnson, Boone, Breach, and
Friedman (2000) argue that a loss of investor confidence and the fall in expected returns on investment
during a financial crisis lead to an increase in expropriation of minority shareholders by controlling
little is known about the restructuring activity during an economic shock and
factors that influence its likelihood.
1 See, for example, John, Lang, and Netter (1992), Ofek (1993), Kang and Shivdasani (1997), and Denis and Kruse (2000). These studies show that following a performance decline, there is a high frequency of corporate restructuring and a significant improvement in post-restructuring operating performance. They also show that corporate governance is closely related to the probability that certain restructuring actions will be taken in response to performance declines. For example, Ofek (1993) shows that high leverage increases the likelihood of asset restructuring and employee layoffs during a performance decline. Kang and Shivdasani (1997) find that the frequency of asset downsizing, layoffs, and removal of outside directors in Japanese firms increases with ownership by the firm’s main bank and other blockholders. Denis and Kruse (2000) also show that asset restructuring is more likely to occur in firms experiencing disciplinary events, such as takeover attempts, shareholder activism, and board dismissals.
2
shareholders.2 Mitton (2002) also argues that, because of an increase in expropriation incentives during a
crisis, corporate governance becomes more critical in explaining cross-firm differences in performance
during the Asian financial crisis. Moreover, financial institutions that experience severe external shocks
might have quite different incentives in dealing with their troubled client firms. For example, as witnessed
by the 2007-2008 global financial crisis, during an economic shock, banks are usually saddled with large
amounts of bad loans. Therefore, during an economic shock, banks are likely to have stronger incentives
to closely monitor their client firms to reduce debt overhang problems. However, it is also possible that
these bad loan problems create incentives for banks to take actions that merely preserve their claims on
borrowers, instead of encouraging borrowers to engage in value maximizing actions (e.g., curtailing
lending to firms with good future prospects, encouraging firms to take inefficient restructuring to generate
cash for the repayment of bank debt, etc.).3
In this study, we use the Korean financial crisis during 1997-1998 as a setting for economic shocks,
and examine how the nature of restructuring activities undertaken by firms experiencing an economic
shock (the “economic shock sample”) is different from that by firms experiencing a large decline in
These arguments suggest that, during an economic shock,
controlling shareholders and other claimholders such as financial institutions have strong incentives to
influence corporate managers to engage in restructuring actions that increase their private benefits.
Therefore, an investigation of restructuring decisions during the economic shock and performance decline
periods will help us better understand the claimholder’s incentives and the role of corporate governance in
different economic environments.
2 Supporting this view, Lemmon and Lins (2003), Claessens, Djankov, Fan, and Lang (2002), and Baek, Kang, and Park (2004) show that during the Asian financial crisis, firms in which controlling shareholders owned more of the voting rights, but fewer cash flow rights, suffered more loss of share values. See also Denis and McConnell (2003) for a review of the literature related to other international corporate governance issues. 3 The following article from the Financial Times illustrates how borrowers are pressured by their banks during the 2007-2008 global financial crisis: “The German BGA exporters' association yesterday forecast a dramatic deterioration in credit conditions in coming months, which would result in massive financing squeeze. Anton Börner, BGA president, told the Financial Times that "for middle- and long-term credit we already have significant difficulties." Even for short-term credit, he expected banks to exert massive pressure on borrowers.” Financial Times (by Ralph Atkins, June 26, 2009, available at http://www.ft.com/cms/s/0/866d2862-61ea-11de-9e03-00144 feabdc0.html).
3
operating performance during a non-crisis period, 1994-1996 (the “performance decline sample”). 4
In examining the determinants of restructuring during a financial crisis, there are at least three reasons
why Korean data are useful in examining the issues raised above. First, Korea experienced a severe
financial crisis in 1997, a crisis that can be characterized by a large drop in the value of its currency and
stock, and a dramatic increase in its interest rates. The exchange rate at the end of October 1997,
approximately one month before Korea sought a rescue package from the IMF, was 902 won per U.S.
dollar. By the end of December 1997, the rate had gone up to 1,836 won per dollar, effectively devaluing
the won by more than 100%. During the same period, the Korea Composite Stock Price Index (KOSPI)
plummeted from 471 to 376 and hit its lowest level of 280 on June 16, 1998. The large depreciation of the
Korean won also subsequently led to a dramatic increase in interest rates. The interest rates on 91-day
commercial paper increased from 13.6% to 21.4% during the last two months of 1997 and peaked at
34.0% on January 7, 1998. To cope with the financial crisis, Korean firms implemented various
Specifically, we examine whether the frequency of restructuring actions, and the factors that affect the
probability of taking such actions are different between these two types of samples. In particular, we
focus on the difference in roles of owner-manager incentives and firm-specific measures of corporate
governance in facilitating restructuring decisions during the economic and performance decline periods.
We use chaebol affiliation and the discrepancy in cash flow rights and voting rights of controlling
shareholders as measures for owner-manager incentives, and leverage (borrowings from banks and
borrowings from non-bank lenders) and equity ownership by unaffiliated financial institutions as
firm-specific measures of corporate governance. We also examine the market’s ex ante valuation of these
effects using the announcement returns of restructuring firms. To the extent that there has been little
empirical evidence on restructurings initiated in response to an economic shock, our analysis should
broaden the understanding of the nature of restructuring activities, and the forces affecting such activities
during different economic environments.
4 Kang, Lee, and Na (2004) examine corporate restructuring in Korea during a performance decline prior to the Asian financial crisis and document that financial structure and chaebol affiliation are important determinants of restructuring.
4
restructuring activities. Therefore, the 1997 Korean financial crisis provides a unique opportunity to
examine restructuring activities conducted in response to an economic shock.
Second, many Korean firms belong to business groups known as chaebols, and ownership of
chaebol-affiliated firms is heavily concentrated in one individual who has almost complete control over all
firms within the group (Bae, Kang, and Kim, 2002; Joh, 2003; Baek, Kang, and Lee, 2006). Even though
the owner-manager of a chaebol firm holds a relatively small portion of its cash flow rights,
cross-holdings among chaebol firms allow each owner-manager to exercise full control over the firm.
This provides controlling shareholders with strong incentives to expropriate minority shareholders, by
investing the firm’s resources to maximize their own or the group’s wealth (La Porta, Lopez-de-Silanes,
and Shleifer, 1999), even when such investments do not maximize the value of the individual firm. These
group and ownership structures allow us to examine how the conflicts of interests between controlling and
minority shareholders influence the restructuring decision of firms during different economic
environments.
Finally, Korean data allow us to examine whether the incentives of large debtholders and unaffiliated
shareholders to monitor their client firms differ during economic shock and performance decline periods.
As documented by Kang (1998), it is well known that many Korean financial institutions, either as
creditors or as shareholders, were passive investors prior to the financial crisis. Furthermore, the Korean
government used banks to exercise control over many firms and, thus, banks’ management was under
great influence from the government. La Porta, Lopez-de Silanes, and Shleifer (2002) show that
government ownership of banks leads to less positive role of banks in emerging markets. However, as
shown in Johnson et al. (2000) and Harvey, Lins, and Roper (2004), the incentives of financial institutions
to monitor their client firms are likely to significantly increase during a financial crisis, because their own
financial difficulties arising from the shock force them to take actions to preserve their claims on these
firms. The Korean data allow us to directly examine these arguments by comparing the disciplinary role
of financial institutions before and during the financial crisis.
We find that chaebol firms in the economic shock sample display an abnormally-high frequency of
restructuring actions, especially cash-generating asset contraction actions (e.g., asset sales), compared to
5
those in the performance decline sample. This is consistent with the view that the economic shock, which
results in a severe credit crunch in the economy and a breakdown of internal capital markets, encourages
chaebol firms to choose the restructuring actions that provide the immediate liquidity that is essential for
survival.
We also find that during the financial crisis, chaebol firms with large discrepancy in cash flow rights
and voting rights of owner-managers are less likely to engage in restructuring actions. To the extent that
restructuring actions are value-enhancing responses to the economic shock and that chaebol
owner-managers have strong incentives to expropriate minority shareholders during the economic shock,
this evidence is consistent with the previous findings that the separation of cash flow rights and voting
rights leads to lower firm values (Johnson et al., 2000; Claessens et al., 2002; Mitton, 2002; Lemmon and
Lins, 2003; Baek, Kang, and Park, 2004). In contrast, the likelihood of restructuring is positively related
to leverage. The effect of leverage on the likelihood of restructuring is more pronounced for chaebol
firms, especially those with large discrepancy in cash flow rights and voting rights of owner-managers,
than for other types of firms. These results are consistent with those of Harvey, Lins, and Roper (2004),
who show that leverage plays a more important monitoring role in firms with higher expected managerial
agency costs.
Moreover, for the economic shock sample, the likelihood of restructuring is positively associated with
equity ownership by unaffiliated financial institutions and the ratio of bank loans to total assets. However,
we do not find any significant role of bank loans for the performance decline sample. These results
suggest that the difficulties of banks and other financial institutions during the period of financial crisis
provide them with strong incentives to actively intervene in firms’ restructuring decisions, in order to
preserve the value of their investments in these firms.
Regarding the market’s reaction to the restructuring announcements, we find that abnormal returns for
the economic shock sample are significantly positive only when restructuring actions generate immediate
cash, or when they are undertaken by chaebol firms. For the performance decline sample, we also find
significant positive returns when restructuring actions generate immediate cash, but do not find such
positive returns when restructuring actions are undertaken by chaebol firms.
6
For the economic shock sample, the results from the multivariate regressions show that the
announcement returns of restructuring actions by chaebol firms with large discrepancies in cash flow
rights and voting rights of owner-managers are lower than those by other types of firms. However, when
these chaebol firms have high leverage, their restructuring actions are viewed more favorably by the
market. These results are consistent with those of previous studies (Claessens et al., 2002; Mitton, 2002;
Lemmon and Lins, 2003; Baek, Kang, and Park, 2004; Harvey, Lins, and Roper, 2004) and suggest that
the market reacts more negatively to restructuring decisions made by firms with controlling shareholders
who have strong incentives to expropriate minority shareholders, whereas it reacts more positively when
better control mechanisms, such as high leverage, reduce such incentives for expropriation. We also find
that abnormal returns for the chaebol restructuring firms are positively related to abnormal returns for the
equally-weighted portfolio of other non-restructuring affiliates in the same group. The positive relation is
stronger if the discrepancy in cash flow rights and voting rights of the controlling shareholders in the
restructuring firm is larger. This finding, together with the result of the negative relation between the
discrepancies in cash flow rights and voting rights of the controlling shareholders and the abnormal return
for the chaebol restructuring firms suggests that such divergence between ownership and control can lead
to wealth transfer from restructuring firms to other member firms in the same group.
Since the conflicts of interests among claimholders during an economic crisis are likely to vary
depending on the institutional settings of the country, we expect the unique institutional environments in
Korea, such as the corporate structure of chaebols, to limit our ability to generalize all the findings in the
paper to other countries. However, our major findings that the 1997 Korean financial crisis creates
different incentives for controlling shareholders, outside shareholders, and creditors concerning
restructuring actions, and that corporate governance plays an instrumental role during the 1997 crisis are
likely to have relevant implications to other countries experiencing a similar economy-wide shock, to the
extent that an economic shock causes similar changes in incentives of various interest groups
involved. Consistent with this view, using a sample of 306 global financial firms across 31 countries,
Erkens, Hung, and Matos (2009) show that ownership structure and corporate governance play an
important role in CEO turnover during the 2007-2008 global financial crisis.
7
This paper proceeds as follows. Section 2 discusses how the restructuring activities and their
announcement effects are related to several measures of owner-manager incentives, corporate governance
mechanisms, and other firm-specific characteristics. Section 3 describes the data and sample
characteristics. In Section 4, we present the results from our empirical analysis. Section 5 summarizes
and concludes the paper.
2. Determinants of restructuring and its announcement effect
In this section we discuss how the factors that affect restructuring decisions play different roles during
an economic shock versus during a performance decline. In many respects, the nature of an economic
shock, such as the 1997 Korean financial crisis, is fundamentally different from that of an individual-level
performance decline. During the period of an economic shock, most firms suffer from financial
difficulties. Since an unexpected economic shock dries up the liquidity in the economy, many firms,
particularly those with high leverage, are likely to run into severe financial trouble. Furthermore, financial
difficulties in the economy increase bad debts in banks’ loan portfolios, forcing banks to curtail lending to
the corporate sector and, thus, creating a severe credit crunch problem. Consequently, their borrowers
must turn to alternative sources of external financing to survive. However, because of capital market
illiquidity and pessimistic investor sentiments during the economic shock, firms with high leverage or
those with highly volatile cash flows tend to experience extreme difficulty obtaining external financing.
Although group-affiliated firms can rely on internal capital markets for funding during a non-crisis period,
these markets do not function well during a crisis period, since other member firms also suffer from the
economic shock. This limits the efficient allocation of capital resources within group-affiliated firms.
Furthermore, since an economic shock has an adverse effect on investor sentiments and employment,
consumer demands tend to significantly drop during the shock period. Therefore, during an economic
shock, many firms not only suffer from existing business losses, they also face few good future investment
opportunities.
In contrast, performance declines typically are caused by firms’ internal or industry-level problems,
such as operational inefficiency, intensive competition, high costs, debt overhang, and so forth. Since the
8
overall financial health of banks and the general condition of capital markets are not likely to be affected
by these problems, firms experiencing a performance decline still can gain access to private and public
financing sources relatively easily, as long as they have good investment opportunities. The operation of
an internal capital market within a business group also enables member firms that are experiencing a
performance decline to easily obtain outside resources. Moreover, since performance declines are specific
to only certain individual companies, investor sentiments and overall employment in the economy are not
likely to be adversely affected by these declines. These differences in economic environments and
difficulties faced by firms during economic and performance shock periods clearly suggest that firms
experiencing a performance decline have fewer incentives to take restructuring activities than those
experiencing an economic shock. They also suggest that the nature of the restructuring process and
owner-manager incentives to undertake restructuring actions are fundamentally different between the two
periods.
Below, we describe how several key variables have a differential effect on restructuring decisions and
restructuring announcement returns during our two sample periods with different economic environments.
• Chaebol. The operation of an internal capital market within the chaebol allows chaebol firms to pool
risks and prop up each other’s performance (Shin and Park, 1999). Thus, during a performance decline,
chaebol-affiliated firms are less likely to engage in restructuring actions than non-chaebol firms are.
However, during the 1997-98 Asian financial crisis, internal capital markets do not function well since
other member firms also suffer from the economic shock. Moreover, Korean banks experienced a severe
credit crunch and were forced to curtail lending to the corporate sector. Consequently, their borrowers,
including chaebol-affiliated firms, had to turn to alternative sources of financing to survive. Therefore,
because of the breakdown of both external and internal capital markets during the economic crisis, even
chaebol firms had difficulty obtaining financing. These scarce resources of chaebols during the financial
crisis suggest that the likelihood of taking restructuring actions by chaebol firms is greater during an
economic shock than during a performance decline. To the extent that the internal capital markets within
the chaebol do not function well during an economic crisis, we also expect that the stock market reacts
9
more favorably when chaebol firms take restructuring actions that relax their liquidity constraints. We
use, as a definition of a chaebol, a dummy variable that equals one if a firm belongs to one of the 30 largest
business groups.5
• Discrepancy between cash flow rights and voting rights. One important characteristic of a chaebol is that
its controlling shareholders have power over all member firms within the chaebol that exceeds their cash
flow rights. To the extent that chaebol owner-managers’ incentives to increase the resource under their
control increase with the divergence between their cash flow rights and voting rights, we would expect
chaebol firms with large ownership discrepancy to engage in restructuring actions less frequently than
their non-chaebol counterparts. This argument also suggests that the substantial discretionary power held
by controlling shareholders in chaebol firms have a negative effect on the market value of firms that
announce restructuring since the controlling shareholders are likely to engage in restructuring for their
own benefits. Moreover, as shown in Johnson et al. (2000) and Lemmon and Lins (2003), if the
owner-managers’ incentives to expropriate minority shareholders increase during a financial crisis, we
would expect any adverse effect of ownership discrepancy on restructuring decisions to be more
pronounced for the economic shock sample than for the performance decline sample. We use the
logarithm of the ratio of voting rights to cash flow rights and a dummy variable that equals one if voting
rights are greater than cash flow rights, as measures of the divergence between cash flow rights and
control rights of the owner-managers.6
• Equity ownership by unaffiliated financial institutions. Although financial institutions in Korea,
including commercial banks, are allowed to invest in stock, they were not very active in monitoring firms
before the Asian financial crisis. This passive role of Korean financial institutions as shareholders
5 The top 30 chaebols have come to represent Korea’s most prominent chaebols during the past three decades. Several previous studies also use top 30 chaebols as a definition of chaebol in their analyses (Joh, 2003; Bae, Kang, and Kim, 2002; Baek, Kang, and Park, 2004). 6 Equity ownership by unlisted affiliated firms also is included in the calculation of voting rights.
10
suggests that they do not play any significant role in restructuring decisions before the Asian financial
crisis. During an economic crisis, however, financial institutions also suffer from the adverse shock, and
experience difficulties selling shares they hold, due to the illiquidity of stock markets. Such difficulties,
coupled with a decline in firm value, may provide financial institutions with strong incentives to monitor
managerial behaviors of the firms, and take actions to preserve the value of their equity investment.
Therefore, equity ownership by unaffiliated financial institutions is expected to be positively associated
with both the likelihood of restructuring and the market value of firms that announce restructuring during
an economic crisis. We measure equity ownership of unaffiliated financial institutions as equity
ownership of commercial banks, merchant banks, and investment trust companies that are not affiliated
with the firms.
• Leverage and bank loans. Jensen (1986) argues that debt-service obligation induces highly leveraged
firms to restructure when their value declines. Consistent with this prediction, Ofek (1993) shows that
high leverage increases the likelihood of asset restructuring and employee layoffs for firms experiencing
performance decline. High leverage also is likely to increase the probability of taking restructuring
actions for firms experiencing an economic shock, because the surge in interest rates and the increase in
bankruptcy risk tend to force managers to make more cautious investment decisions. These arguments
suggest that the probability of taking restructuring actions and the valuation effect of restructuring events
are positively associated with leverage. We measure leverage as the ratio of total debt to total assets.
Among the different types of debt claims, bank loans likely play an important role in corporate
restructuring. Korean firms traditionally have been heavily reliant on bank financing, maintaining close
and long-term ties with their banks. Therefore, as informed monitors, banks are expected to play an
important role in corporate restructuring, in response to a performance decline (Fama, 1985).7
7 Kang (1998), however, argues that, during the normal period, Korean banks have few incentives to play the role of an active monitor, especially for chaebol firms, because most bank loans to chaebol firms are guaranteed by the cross-debt guarantees among them.
This role
of banks in corporate restructuring is expected to be manifested particularly when there is a shock to the
11
economy. When banks experience financial difficulties due to the adverse shock, they will have strong
incentives to actively monitor their client firms, in order to preserve their debt claims on these firms.
These arguments suggest that a higher fraction of bank debt in a firm’s capital structure increases the
likelihood of taking restructuring actions during an economic shock to a greater extent than during a
performance decline. They also suggest that the effect of bank loans on the market value of firms that
announce restructuring is more positive during an economic shock than during a performance decline. We
use the ratio of bank debt to total assets to measure the strength of a firm’s ties to a bank.
• Other control variables. In addition to the variables discussed above, we also consider other control
variables in examining the likelihood of restructuring decisions and valuation effects, including firm size
(the logarithm of the market value of equity), investment opportunities (Tobin’s q as measured by the ratio
of the market value of equity plus the book value of debt to the book value of total assets), liquidity (the
ratio of liquid assets (cash plus marketable securities) to total assets) and cash flows (the ratio of cash flow
(operating income plus depreciation) to total assets), and an export ratio (the ratio of exports to sales).
3. Data and summary statistics
Our sample of firms that experienced an economic shock consists of 580 nonfinancial firms listed on
the Korean Stock Exchange (KSE) that were not under the supervision of the bankruptcy court as of
November 18, 1997. We exclude the firms under court supervision, because important strategic decisions
for these firms, such as restructuring, tend to be made by courts, not by management. To examine the
restructuring activities of these firms, we use the Korea Investor’s Network for Disclosure System
(KINDS) and the Korean Economic Daily Newspaper. 8
8 Korean listed firms are required to disclose their asset sales immediately whenever the amount of the sale is greater than 10% of the value of their total assets. This disclosed information is readily available to the public at the KINDS. Thus, our sample collection procedures that use both the KINDS and a financial newspaper are less likely to be subject to a selection bias, compared to those that use only financial newspapers. However, to ensure the robustness of our results, we also obtain information about restructuring activities for our sample firms from the firm-level financial data and re-estimate the regressions reported in the following section. Specifically, using the Listed Company Database of the Korean Listed Companies Association, we calculate the changes in tangible assets and
All sources are examined for articles on
12
restructuring activities during the one-year period from November 18, 1997 to November 17, 1998.
November 18, 1997 is the date when the IMF proposed a rescue package to Korea, to help it overcome the
financial crisis that had started in the middle of that month. During that one-year period, starting
November 18, 1997, the stock market plunged to its lowest level ever, on June 16, 1998, recording the
lowest KOSPI level of 280, and many Korean firms implemented various restructuring activities and
reforms to overcome the financial crisis. We identify each firm’s chaebol affiliation using the Korea Fair
Trading Commission’s Annual Statistics, and find that 131 out of the 580 firms belong to one of the top 30
business groups.
To more closely examine the nature of restructuring activities during the economic shock, we
construct a comparison sample of firms experiencing a performance decline during the 1994-1996 period,
and compare the restructuring activities between the economic shock and performance decline samples.
We start the sample period from 1994 because, before 1994, detailed data on controlling ownership are not
available. To identify comparison firms, we follow the methodology of Kang and Shivdasani (1997).
Specifically, for each year, we choose, as comparison firms, those that had experienced more than a 50%
drop in their operating income (i.e., ROA) during the fiscal year,9
Since the magnitude of the economic shock cannot be directly comparable to the magnitude of a
performance shock, and since the economic shock sample includes all non-financial firms during
1997-1998, whereas the performance decline sample includes only firms experiencing poor operating
performance, it is not possible to identify a control firm that has the same characteristics as the firm
experiencing economic shock. Thus, the purpose of identifying a comparison firm is not to directly
but were not under the supervision of
the bankruptcy court.
employment from the fiscal year end of 1996 to the fiscal year end of 1998. We then consider that a firm takes a restructuring action if its industry-mean-adjusted change in tangible assets (or employment) is in the bottom quartile of the sample. The untabulated tests show that the results are qualitatively similar to those reported in the paper. 9 Unlike Kang and Shivdasani (1997), who restrict their sample firms to those that were initially healthy before their performance declines, we do not impose “healthy” conditions when we select our sample firms. This is because the economic shock sample includes all non-financial firms listed on the KSE, irrespective of whether they initially were healthy before the shock. Therefore, imposing healthy conditions only on the performance shock sample could cause a selection bias problem, and make the comparison less meaningful.
13
compare its restructuring actions with those of the firm experiencing economic shock with similar
characteristics. Rather, our purpose is to provide a reference point to aid in interpreting the results for the
economic shock sample. Since evidence on the restructuring activities responding to performance
declines is well documented for U.S. and Japanese firms, this evidence, together with the results derived
from restructuring actions by Korean firms during a performance decline, is expected to provide important
guidelines for evaluating the restructuring actions in response to an economic shock.
To identify the restructuring activities of firms in the performance decline sample, we search the
KINDS and the Korean Economic Daily Newspaper for the one-year period from six months before to six
months after the fiscal year end of the poor performance year. For example, if the fiscal year end of a firm
is December 1994, we search sources from July 1994 to June 1995. As an alternative search period, we
identify the restructuring activities of firms during the year of poor performance. Although not reported
here, we examine the robustness of our findings by repeating all the analyses presented below using this
new set of restructuring activities, and obtain results that are qualitatively similar to those reported in the
paper. There are 193 firms in the performance decline sample. Out of these 193 sample firms, 33 belong
to one of the top 30 chaebol groups.
We obtain daily stock return data from the Korea Investors Service-Stock Market Analysis Tool
(KIS-SMAT), which includes the returns of all firms listed on the KSE. Financial data and bank loan data
are obtained from the Listed Company Database of the Korean Listed Companies Association and annual
reports, respectively. Ownership information of unlisted affiliated firms is obtained from the on-line
database of Korea Information Service, KIS-Line.
In Table 1, we present the summary statistics for ownership structure and other financial
characteristics of our sample firms. We divide the sample firms into chaebol and non-chaebol firms.
Following La Porta, Lopez-de-Silanes, and Shleifer (1999), Claessens, Djankov, and Lang (2000), Mitton
(2002), and Lemmon and Lins (2003), we measure the cash flow rights of the owner-managers
(controlling shareholders) as the sum of the direct equity ownership of the owner-managers and their
family and the product of the ownership stakes obtained indirectly along the chain in a pyramid structure
(by tracing up to two layers of control chains). Voting rights are taken as the sum of direct equity
14
ownership of the owner-managers and their family and equity ownership by affiliated firms. It is widely
recognized that member firms within chaebols are connected by extensive reciprocal shareholding
agreements, and that chaebol owner-managers exercise substantial control over affiliated firms.
Therefore, in accordance with Lemmon and Lins (2003) and Harvey, Lins, and Roper (2004), we include
equity ownership by affiliated firms in estimating voting rights of the owner-managers. For the economic
shock sample, the mean cash flow rights and voting rights of controlling shareholders are significantly
lower for chaebol firms (11.92% and 25.34%) than those for non-chaebol firms (27.94% and 30.74%) at
the 0.01 level. However, as expected, the mean ratio of voting rights to cash flow rights of chaebol firms
is significantly greater than that of the non-chaebol firms (10.67 versus 1.35). The mean equity ownership
by unaffiliated financial investors also is significantly larger for chaebol firms than for non-chaebol firms
(14.12% versus 6.80%). The medians show a similar pattern.
Leverage is significantly different between the two groups, as well. For the economic shock sample,
the mean leverage ratios for chaebol firms and non-chaebol firms are 76.4% and 67.0%, respectively. The
average firm borrows about 31.2% of its total assets from banks. The mean ratios of bank loans to total
assets are similar for chaebol firms and non-chaebol firms (30.8% versus 31.3%).
For the economic shock sample, we also find that chaebol firms are larger than non-chaebol firms.
The average firm size of chaebol firms, measured by the market value of equity, is more than four times
that of non-chaebol firms. The difference in firm size is statistically significant at the 0.01 level. The
mean Tobin’s q is lower for chaebol firms than for non-chaebol firms (1.01 versus 1.11). For chaebol and
non-chaebol firms, the mean ratios of cash flow to total assets are, respectively, 5.7% and 5.5%, and the
mean ratios of liquid assets to total assets are 5.2% and 10.9%. Although the mean difference in cash flow
ratios between chaebol and non-chaebol firms is not significant, the difference in liquidity ratios is
significant at the 0.01 level. The average ratio of exports to sales is 27.6% for chaebol firms and 24.7% for
non-chaebol firms, though the difference is not significant.
For the performance decline sample, the results show similar patterns to those for the economic shock
sample, except that the voting rights of controlling shareholders and Tobin’s q are not significantly
different between the chaebol and non-chaebol firms.
15
4. Empirical results
4.1. Restructuring activities
In this section, we examine various corporate restructuring activities that our sample firms engage in,
in response to the economic shock and a performance decline. Following Denis and Kruse (2000) and
Kang and Shivdasani (1997), we classify the restructuring activities into asset contraction actions and
employment changes. Asset contraction actions include the sale of assets, investment cuts, plant closures,
product or business line withdrawals, and spin-offs of units. Employment changes include layoffs,
reductions in compensation or bonuses, and other actions that significantly affect the composition or
compensation of a firm’s employees, such as lowering the mandatory retirement age and offering early
retirement incentives.
Table 2 shows the frequency of restructuring actions for both the economic shock and the
performance decline samples. Out of 580 sample firms experiencing the economic shock, 195 (34%)
engage in restructuring actions. In comparison, within a sample of 193 firms that have experienced a
performance decline, 51 (26%) engage in restructuring actions. The tests of difference in the frequency of
firms engaging in restructuring actions show that firms in the economic shock sample are significantly
more likely to engage in such actions than are those in the performance decline sample. The difference is
particularly evident among chaebol firms. For example, while more than 67% of chaebol firms engage in
restructuring actions during the economic shock, the corresponding number during a performance decline
is only 42%. The difference in the frequency of chaebol firms that undertake restructuring actions
between the economic and the performance decline samples is significant at the 0.01 level. In contrast, the
frequency of restructuring actions by non-chaebol firms during the economic shock is not significantly
different from that during a performance decline (24% versus 23%).
These results indicate that chaebol firms in the economic shock sample display an abnormally-high
frequency of restructuring actions compared to those in the performance decline sample, suggesting that
the poor functioning of internal capital markets within a chaebol during the 1997-98 financial crisis had a
significant effect on restructuring incentives of its member firms.
16
Among various restructuring activities, asset contraction actions occur more frequently than
employment changes, within both the economic shock and the performance decline samples. During the
economic shock (performance decline), 31% (25%) of the sample firms undertake asset contraction
actions, whereas only 8% (3%) of the sample firms undertake employment changes. In comparison, using
a sample of 92 Japanese and 114 U.S. firms that experienced a decline of at least 50% in their operating
income in a given year during 1986 and 1990, Kang and Shivdasani (1997) show that 23% of their sample
Japanese firms and 49% of their U.S. sample firms engage in some form of asset downsizing actions
during the year of the performance shock and the following year. Ofek (1993) finds that 23% of his U.S.
sample firms announce asset restructuring during the distress year, and Denis and Kruse (2000) show that
27% of their U.S. sample firms that experience performance declines undertake asset restructuring in the
year of a performance decline. Thus, Korean firms that experience an economic shock are more likely to
engage in asset contraction policies than are Japanese firms experiencing performance declines, but
appear to undertake asset contraction actions as frequently as U.S. firms do.
Of the various asset contraction actions, the sale of assets is the most frequently observed response
among firms in both the economic shock and performance decline samples. About 26% (21%) of the
firms in our economic shock (performance decline) sample engage in asset sales. In comparison, Kang
and Shivdasani (1997) find that 4% of their Japanese sample firms are involved in asset sales. The
corresponding number in the Denis and Kruse’s (2000) sample of U.S. firms is 15%. The higher
frequency of asset sales in Korean firms that experience the economic shock than for Japanese, Korean,
and U.S. firms that experience performance declines suggests that, to relax their liquidity constraints,
firms experiencing an economic shock prefer to choose the restructuring type that generates immediate
cash.
To further investigate this issue, we divide asset downsizing actions into two categories, based upon
whether or not these actions generate cash inflows. The results show that the majority of asset downsizing
actions in the economic shock sample is associated with cash generation. For example, 55% of chaebol
firms in the economic shock sample engage in cash-generating asset contraction actions, while only 19%
engage in non-cash-generating asset contraction actions (e.g., withdrawal from the line of business,
17
discontinuing or consolidating operations, and closing plants or headquarters). The corresponding
numbers for chaebol firms in the performance decline sample are only 30% and 12%, respectively.
Regarding changes in employment and compensation policies, as shown above, about 8% of the firms
in the economic shock sample and 3% of the firms in the performance decline sample change their
employment and compensation policies. In comparison, Ofek (1993) and Denis and Kruse (2000)
document an employee-layoff occurrence rate of 28% and 13% for their samples of poorly performing
U.S. firms, respectively, and Kang and Shivdasani (1997) report that 28% of poorly performing Japanese
firms undertake employment change policies during the performance decline year and the year subsequent
to it. Thus, Korean firms in the economic shock sample appear to display an abnormally high frequency of
employee layoffs, compared to Korean firms in the performance decline sample, but show a low
frequency compared to U.S. and Japanese firms that experience a performance decline. We also find that,
while the frequency of adopting changes in employment and compensation policies during an economic
crisis is significantly higher for chaebol firms than for non-chaebol firms (21% versus 4%), such a
significant difference does not exist during a performance decline (6% versus 2%).
In summary, compared to the firms experiencing a large decline in operating performance, those
experiencing the economic shock tend to engage more frequently in restructuring actions. The difference
is more evident among chaebol firms. These results are consistent with the view that the capital market
illiquidity and the lack of profitable investment opportunities observed during an economic crisis make
firms, especially chaebol firms, financially-constrained and, therefore, induce them to choose contraction
policies that can preserve liquidity.
4.2. Determinants of restructuring decisions
In this section, to clarify the relation between the key variables of interest discussed in Section 2 and
the likelihood of restructuring activity, we use multivariate logit regressions in which the dependent
variable is set to one if a particular restructuring action occurs, and zero otherwise. Explanatory variables
are measured at the fiscal year-end that immediately precedes the economic shock and a performance
decline.
18
Table 3 examines the factors that influence the likelihood of restructuring in response to an economic
shock. All variables that are interacted with a dummy variable are demeaned by subtracting the mean of
sample firms from their original values. In model (1), we use, as explanatory variables, chaebol
affiliation, the logarithm of the ratio of voting rights to cash flow rights as a measure of ownership
discrepancy, interaction term between these two variables, leverage, and equity ownership by unaffiliated
financial institutions. In accordance with Lemmon and Lins (2003), we winsorize the logarithm of the
ratio of voting rights to cash flow rights at the 95th percentile, to avoid overemphasizing firms with
extreme ownership discrepancy. The logit model also controls for firm size, Tobin’s q, the ratio of cash
flow to total assets, the ratio of liquid assets to total assets, the ratio of export to sales, and industry
dummies. We find that the coefficient on the chaebol dummy is positive and significant at the 0.05 level,
indicating that chaebol firms are more likely to engage in restructuring actions than non-chaebol firms.
However, the coefficient on the interaction term between the ownership discrepancy and the chaebol
dummy is negative and significant. Thus, chaebol firms are less likely to engage in restructuring actions
when their controlling shareholders have large discretionary power relative to their cash flow rights. The
coefficient on leverage is positive and significant at the 0.01 level. Thus, controlling for other factors,
firms with high leverage are more likely to contract when the economy is adversely affected. This result is
consistent with findings by Ofek (1993) and supports Jensen’s (1989) argument that high leverage forces
firms to respond quickly to a decline in value. We also find that unaffiliated financial institutions perform
an important role by increasing the likelihood of asset contractions during the financial crisis. This result
is consistent with theoretical work on the role of outside shareholders by Demsetz and Lehn (1985) and
Shleifer and Vishny (1986). Among control variables, the coefficient on firm size is positive and
significant whereas the coefficient on Tobin’s q is negative and significant at the 0.01 level. These results
suggest that larger firms and firms with fewer investment opportunities are more likely to undertake
restructuring actions.
To examine whether the effect of leverage on the likelihood of restructuring is different between
chaebol and non-chaebol firms, in model (2) we interact the leverage ratio with a chaebol dummy. During
an economic shock, the resources available for chaebol firms are likely to be limited, because the internal
19
capital market does not function well, and a liquidity problem in one member firm can spread into the
whole group because of cross-debt guarantees within chaebol firms. Moreover, given that chaebol firms
typically maintained higher leverages than non-chaebol firms during our sample period (Bae, Kang, and
Kim, 2002), chaebol firms are likely to be more vulnerable to an economic shock. Therefore, chaebol
firms with high leverage are likely to take more frequent restructuring actions than their non-chaebol
counterparts. Consistent with this view, we find that the coefficient on the interaction term is positive and
significant. However, the positive coefficient on the chaebol dummy in the previous regression becomes
insignificant. These results indicate that the high probability of restructuring by chaebol firms reported in
model (1) is mainly driven by those firms with a high leverage ratio. Thus, during an economic shock, the
disciplinary role of debt is particularly strong for highly-leveraged chaebol firms.
In model (3), we disaggregate the leverage into the ratio of bank loans to total assets and the ratio of
non-bank debt (total debt – bank loans) to total assets, and include the interaction term between the
chaebol dummy and two leverage variables. The coefficients on both bank debt and non-bank debt ratios
are positive and significant, but the coefficient on non-bank debt ratio is smaller and less significant than
the coefficient on bank debt ratio. These results suggest that firms are more likely to engage in
restructuring when their banks hold a significant debt claim, supporting the hypothesis that banks with
large debt claims have strong incentives to monitor a firm. The importance of bank debt in corporate
restructuring is consistent with the results in Gilson, John, and Lang (1990), who show that U.S. firms
with large bank borrowings are more likely to restructure their debt. The results, however, also are
consistent with the view that banks become active monitors during an economic shock, because of their
own financial difficulties and incentives to preserve their investments in client firms. Model (3) also
shows that the coefficients on both interaction terms – between the chaebol dummy and the bank loan
ratio, and between the chaebol dummy and the non-bank debt ratio – are positive and significant, but the
coefficient on the former interaction term is larger and more significant.
Models (4) and (5) replace the chaebol dummy in models (2) and (3) with a dummy variable that
equals one if the ratio of voting rights to cash flow rights of owner-managers for chaebol firms is greater
than one, and zero otherwise. We expect that the control function of leverage to facilitate restructuring
20
decisions is more pronounced for chaebol firms with large discrepancy in cash flow rights and control
rights of owner-managers. For example, Harvey, Lins, and Roper (2004) show that leverage plays an
important monitoring role in firms with high expected managerial agency costs, and that debt creates
value when there is a large degree of separation between cash flow rights and control rights. Thus, during
an economic crisis, leverage is expected to increase the likelihood of restructuring, particularly for
chaebol firms with large ownership discrepancy. Consistent with this view, we find that the coefficients
on the interaction term between a dummy variable for chaebol ownership discrepancy and the leverage
ratio and the interaction term between this dummy and the bank loan ratio are positive, but only the
coefficient on the latter interaction term is significant.
In untabulated tests, we also examine the likelihood of restructuring in response to a performance
decline. Similar to the findings for the economic shock sample, the coefficient on the leverage ratio is
positive and significant for the performance decline sample. However, when we disaggregate the leverage
ratio into the bank loan and non-bank debt ratios, the coefficient on the bank loan ratio becomes
insignificant, whereas the coefficient on the non-bank debt ratio turns out to be positive and significant.
The coefficient on equity ownership by financial institutions is also insignificant. These results suggest
that, prior to the financial crisis, banks and unaffiliated financial institutions, either as debtholders or as
shareholders, do not have incentives to monitor corporate managers, supporting the notion that Korean
financial institutions played a passive role before the financial crisis, as discussed in Section 2. We also
find that, unlike the results observed with the economic shock sample, the coefficient on the ownership
discrepancy variable and the coefficient on its interaction with the leverage ratio are not significant. The
coefficients on most control variables are also insignificant.10
10 To examine whether the impacts of ownership discrepancy and leverage ratio on a firm’s restructuring decision during an economic shock are different from those during a performance decline, in untabulated tests, we estimate the logit regressions for a pooled sample of restructuring actions during economic shock and performance decline periods. We find that chaebol firms with ownership discrepancy in the economic shock sample are less likely to restructure than non-chaebol counterparts in the economic shock and performance decline samples. We also find that compared to non-chaebol firms in the economic shock and performance decline samples, highly-leveraged chaebol firms experiencing an economic shock are more likely to engage in restructuring actions. This positive effect of leverage ratio on restructuring actions during the economic crisis is particularly strong when chaebol firms with
21
Overall, the analysis of restructuring actions for the economic shock sample indicates that the
probability of taking restructuring actions increases when firms maintain high leverage, particularly bank
debt, in their capital structure. In contrast, chaebol firms with large ownership discrepancy are less likely
to restructure in response to an economic shock, possibly due to the owner-managers’ distorted incentives
to maximize the group size. We also find that the probability of taking restructuring actions increases
when unaffiliated financial institutions that are less likely to be influenced by managers hold large equity
claims, and when chaebol firms, especially those with ownership discrepancies, have high leverage or
maintain close lending relationships with banks.
4.3. The valuation effects of restructuring events
Previous studies show that restructuring activities caused by performance shocks lead to an increase in
firm value (Jain, 1985; Denis and Kruse, 2000). In this section, we examine the valuation effects of
restructuring announcements during an economic shock and a performance decline. Restructuring
announcements, in response to an adverse economic shock and a large decline in operating performance,
will be value-enhancing events, if firms sell unprofitable assets or operations, reduce labor expenses, or
strengthen their strategic position by focusing on core businesses. In particular, during an economic
shock, restructuring actions that generate immediate cash inflow might be greeted more favorably by
investors since such actions increase a firm’s liquidity, which is crucial in coping with a financial crisis.
To assess the valuation effect of corporate restructuring, we use a standard event-study approach. We
identify initial public announcement dates of the restructuring activities from the KINDS, and the Korean
Economic Daily. As the announcement date, we use the date that a news announcement first appeared in
either of these two publications.
For each event, we compute the abnormal return using the market model. We use the KOSPI return as
the market return proxy. We obtain our estimates of the market model by using 200 trading days of return
data, beginning 220 days before and ending 21 days before the restructuring announcement. We then
ownership discrepancy have a high bank loan ratio, indicating that the disciplinary role of bank loans becomes stronger when the agency problem of controlling shareholders is severe (Harvey, Lins, and Roper, 2004).
22
calculate daily abnormal returns and sum the daily abnormal returns over three days around the
restructuring announcement date, to obtain the cumulative abnormal return (CAR (-1, 1)). To avoid biases
in the estimation of CAR, we exclude observations if the firm makes other restructuring announcements
during the estimation period of the market model. We use the t-statistic to test the hypothesis that the
average CARs are equal to zero, and the Wilcoxon sign-rank test statistic to test the hypothesis that the
CARs are symmetrically distributed around zero.
Panels A and B of Table 4 report the CARs (-1, 1) of restructuring actions for the economic shock
sample and for the performance decline sample, respectively. During the economic shock, the chaebol
firms that announce restructuring policies realize significantly positive returns. The mean and median
CARs (-1, 1) for chaebol firms are 3.4% and 2.5%, respectively, both of which are significant at the 0.01
level. In contrast, the mean and median CARs (-1, 1) for non-chaebol firms are not significant. Separating
restructuring actions into cash-generating and non-cash-generating actions, we find that, for chaebol
firms, cash-generating actions lead to positive and significant mean and median returns (4.6% and 2.6%)
and non-cash-generating actions lead to negative and insignificant mean and median returns. The
difference in mean CARs (-1, 1) is significant at the 0.05 level. We find similar patterns for non-chaebol
firms. However, for non-chaebol firms, the mean CAR (-1, 1) for non-cash-generating actions is negative
and marginally significant (-3.7%), but the mean CAR (-1, 1) for cash generating actions is positive and
insignificant (0.5%). These results are generally consistent with those of Denis and Kruse (2000) who
show that abnormal returns for asset sale announcements during performance declines are positive and
significant, whereas those for cost-cutting or layoff announcements are negative and insignificant. Given
that, during an economic crisis, most firms suffer from the liquidity shock and that the internal capital
markets within the chaebol do not function well, our results suggest that investors are disappointed when
firms do not solve their liquidity problems by taking actions that immediately generate cash inflow.
However, as shown in Panel B of Table 4, we do not find any significant valuation effects for restructuring
actions during a performance decline, except for a marginally positive CAR (-1, 1) for cash-generating
restructuring actions of the total sample.
23
To examine whether the announcement returns of restructuring actions are related to owner-manager
incentives and firm-specific measures of corporate governance, we present the estimates from
multivariate regressions using CARs (-1, 1) as the dependent variables. Since the announcement returns
differ significantly depending on whether the actions generate cash flows, we run the regression models
for cash-generating and non-cash-generating actions separately. Table 5 reports the regression estimates
for the sample of cash-generating restructuring actions during the economic shock. All variables that are
interacted with a dummy variable are again demeaned by subtracting the mean of sample firms from their
original values.
In model (1), the coefficient on the interaction term between the chaebol dummy and the logarithm of
the ratio of voting rights to cash flow rights is negative and significant. This result is consistent with the
view that the separation of the cash flow rights and control rights of chaebol owner-managers increases
their incentives to expropriate other investors during an economic shock. In contrast, model (2) shows
that the market responds more positively when restructuring activities are announced by chaebol firms
with a higher leverage ratio. Model (3) further shows that the positive role of leverage in model (2) is
more pronounced for chaebol firms with ownership discrepancy. This result suggests that debt disciplines
chaebol owner-managers with severe agency problems, and influences them to follow value-enhancing
restructuring decisions. Model (4) demonstrates that the coefficients on both interaction terms between
bank debt and a dummy variable for chaebol firms with ownership discrepancy and between non-bank
debt and a dummy variable for chaebol firms with ownership discrepancy are positive and significant. In
contrast, for the subsample of non-cash-generating restructuring actions during an economic shock, none
of the variables is significant (not reported).11
11 To examine whether the impacts of ownership discrepancy and leverage ratio on CARs (-1,1) during an economic shock are different from those during a performance decline, in untabulated tests, we estimate the regressions for a pooled sample of restructuring actions during economic shock and performance decline periods. We find that restructuring actions by chaebol firms with ownership discrepancy in the economic shock sample result in more negative announcement returns than do those by non-chaebol firms in the pooled sample. We also find that the positive effects of leverage on firm value are significantly higher for chaebol firms, particularly for chaebol firms with ownership discrepancies, in the economic shock sample than for non-chaebol firms in the pooled sample.
24
Overall, our results show that, during an economic shock, the announcement effects of
cash-generating restructuring actions undertaken by chaebol firms are positive and significant. In
contrast, the announcements of restructuring by chaebol firms with divergence in cash flow rights and
control rights of controlling shareholders and those with a lower leverage ratio are greeted more
negatively by investors. These results suggest that chaebol owner-managers have different objectives
than their non-chaebol counterparts when they restructure their firms, especially during an economic
shock, and that the stock market pays attention to these differences in managerial incentives.
4.4. Effect of restructuring actions on the equally-weighted portfolio of other member firms
If a firm’s restructuring decisions during an economic shock are motivated by controlling
shareholders’ incentives to maximize their own private benefits or by incentives to share resources among
member firms, restructuring announcements will have value implications not only for restructuring firms
but also for other member firms in the same chaebol. For example, if a chaebol firm engages in
cash-generating restructuring to increase controlling shareholders’ private benefits and if such
restructuring sends bad signals to the market about the quality of the governance structures the chaebol
adopts, it could have a negative impact on the value of other group affiliates. On the other hand, if cash
generated through a firm’s restructuring implies more resources available for intra-group propping, such
restructuring announcements will have a positive effect on the market value of other member firms in the
same group.
To address this issue, we estimate the equally-weighted portfolio returns of other firms that belong to
the same chaebol as the restructuring firm. Abnormal returns for other group affiliates are calculated by
using a portfolio approach. The market-model parameters are estimated using returns of the
equally-weighted portfolio of other firms in the same group and then used to calculate the daily abnormal
returns of the portfolio. The daily abnormal returns are accumulated to obtain the portfolio CAR from day
-t to day +t.
Panel A of Table 6 shows the mean and median equally-weighted portfolio CARs (-1, 1) of other
member firms around restructuring announcement dates. For the subsample of cash generating
25
restructuring, both the mean and median equally-weighted portfolio CARs (-1,1) of other member firms
are positive and significant at the 0.01 level. These results are consistent with the propping view of
restructuring during an economic shock. However, for the subsample of non-cash generating
restructuring, we do not find any significant abnormal returns for other member firms.
Panel B of Table 6 reports the results from regressing the equally-weighted portfolio CAR (-1, 1) of
other member firms on restructuring firms’ CARs (-1, 1) and financial characteristics. Model (1) shows
that the equally-weighted portfolio returns of other member firms are positively and significantly related
to the returns of restructuring firms. To the extent that resources generated through restructuring are an
important indicator of the chaebol’s ability to prop up its member firms, this result is consistent with the
propping view regarding business groups. Model (2) shows that the coefficient on the interaction term
between the CARs (-1, 1) of restructuring firms and the dummy for high cash flow rights of the controlling
shareholders (equals 1 if the cash flow right is above the sample median) is negative and significant. Thus,
the sensitivity of abnormal returns for the equally-weighted portfolio of non-restructuring affiliates to
abnormal returns for the restructuring firms is lower when the controlling shareholders have high cash
flow rights in the restructuring firm. To the extent that controlling shareholders with high cash flow rights
in a certain firm have a strong incentive to discourage it from engaging in restructuring to prop up other
member firms, this result further supports the propping view for business groups during an economic
shock period. In contrast, model (4) shows that the positive association between abnormal returns for the
restructuring firms and abnormal returns for the equally-weighted portfolio of non- restructuring affiliates
is stronger if the discrepancy in cash flow rights and voting rights of the controlling shareholders in the
restructuring firm is larger. To the extent that chaebol owner-managers’ power to reallocate the resources
among member firms increases with their ownership discrepancy in restructuring firms, this result
suggests that the propping effect is stronger when the controlling shareholders have concentrated
ownership in the firm that exceeds their cash flow rights and have excessive power over the firm.
Overall, the results in Table 6 suggest that the market perceives propping-motivated restructuring activities by a
chaebol-affiliated firm during an economic shock as group-wide events. They also suggests that the market’s ex ante
valuation of propping effects is significantly affected by cash flow rights and discrepancy in cash flow rights and
26
voting rights of the controlling shareholders in the restructuring firm.12
4.5. Additional tests
In this section we briefly discuss the results of additional tests regarding the probability of taking
restructuring actions during an economic shock, valuation effects of restructuring, and the effect of
restructuring actions on the long-term operating performance. For the sake of brevity, we do not report the
results in the table.
• Marginal effects of key determinants of downsizing actions
To examine whether potential bias in interpreting logit regressions with interaction terms also exists in
our data and to determine whether the key determinants of restructuring policies that are statistically
significant in Table 3 are also economically significant, we separately calculate the marginal effects of key
variables for chaebol and non-chaebol firms, as suggested by Powers (2005). For the subsample of
chaebol firms during the economic shock, the marginal effect coefficient on the dummy for ownership
discrepancy is a significant -0.153, suggesting that the probability of taking restructuring actions by
chaebol firms with ownership discrepancy is about 15.3% lower than that by chaebol firms without
ownership discrepancy. We also find that for chaebol firms a one-standard deviation increase in
ownership by unaffiliated financial institutions (9.3%) from the sample mean (14.1%) would result in a
13.5% increase in the likelihood of taking a restructuring action. Similarly, a one-standard deviation
increase in the leverage ratio (11.4%) or bank loan ratio (14.9%) from the sample mean is associated with
a 16.0% or a 25.9% increase in the likelihood of restructuring, respectively. In contrast, among these
variables only leverage and bank loan ratios are statistically significant for the subsample of non-chaebol
firms during the economic shock with smaller marginal effects. For the subsample of chaebol firms
during a performance decline, none of these variables are significant.
12 In unreported tests, we also experiment with the value-weighted portfolio CAR and obtain the results that are qualitatively similar to those reported in the paper.
27
• Quantifying the size of restructuring actions
In this paper, we have measured the restructuring action using a dummy variable. To check the
robustness of our results, we quantify the size of restructuring and use it as a measure of the extent of
restructuring. Of 156 firms that undertake cash-generating downsizing actions, we are able to collect the
data on the size of restructuring and other relevant information for 114 firms. Using these 114
restructuring firms, together with non-restructuring firms, we reestimate the regressions in Table 3.
Specifically, we use as a dependent variable the ratio of cash generated from restructuring to the market
value of equity and as independent variables the same explanatory variables as those used in Table 3. We
obtain qualitatively similar results.
• Fire sales of assets
It is possible that firms destroy values by being forced to engage in fire sales to solve short term
liquidity problems. In particular, it is possible that non-chaebol firms under liquidity constraints are more
likely to engage in fire sales than chaebol firms that can rely on internal capital markets for funding. To
address this issue, we have collected the data on prices and book values of assets sold from the KINDS and
the Korean Economic Daily Newspaper. Although 151 sample firms have conducted 221 asset sales
during the financial crisis period, both prices and book values are available only for 93 cases, 48 (45) of
which are undertaken by chaebol (non-chaebol) firms. Although this small sample size prohibits us from
conducting a meaningful analysis, we examine whether the market’s responses to assets sales vary
depending on the ratio of price-to-book value of assets sold. We also examine whether the ratio of
price-to-book value of assets sold is significantly different between chaebol firms and non-chaebol firms.
Price-to-book value ratios are likely to be lower for those assets sold at fire-sales prices, holding
everything else constant. In untabulated tests, we find that there is no significant difference in the
price-to-book value ratio between chaebol and non-chaebol firms. When we use the price-to-book value
ratio as a dependent variable and a dummy variable for non-chaebol firms together with other control
variables as the independent variables, none of the coefficients are statistically significant. Finally, we
find that the price-to-book value ratio is positively and significantly related to CARs (-1, 1), but the
28
interaction term between this ratio and the non-chaebol dummy is not significantly related to CARs (-1, 1).
These results suggest that the extent of fire sales of assets during an economic crisis is similar between
chaebol and non-chaebol firms and the effect of such fire sales on firm value is statistically
indistinguishable between these two groups of firms, possibly due to the breakdown of internal capital
markets within the chaebol.
• Other determinants of downsizing actions
Block ownership: Since Mitton (2002) shows that firms with concentrated ownership suffered less during
the Asian financial crisis, we first examine whether large block ownership has an effect on a firm’s
restructuring decisions. We measure the ownership concentration by the sum of block holdings by all
shareholders who own 5% or more of issued shares (summed blockholder concentration) and include it as
an additional explanatory variable for the regressions in Table 3. We find that the coefficients on summed
blockholder concentration are not significant. However, when we subdivide summed blockholder
concentration into affiliated (block holdings by owner-managers, their family members, and affiliated
firms) and unaffiliated ownership, and interact these variables with the chaebol dummy, the interaction
variable between affiliated ownership and chaebol dummy is significantly and negatively related to the
likelihood of restructuring. These results are consistent with those of Baek, Kang, and Park (2004), who
show that, during the Asian financial crisis, higher equity ownership by insiders and affiliated firms had a
greater adverse effect on firm values when the firm belonged to one of the top 30 chaebols.
Foreign debt versus domestic debt: Allayannis, Brown, and Klapper (2003) show that, during the Asian
financial crisis, the firms that used the foreign currency denominated debt and hedged it to local currency,
experienced the largest drop in their market value. To examine whether the currency of denomination of
debt matters in determining restructuring decisions, we divide total leverage used in Table 3 into foreign
and domestic borrowings. We find that only the ratio of domestic debt to total assets is significantly and
positively related to the likelihood of taking restructuring actions.13
13 The potential explanation for the insignificance of foreign debt in the regressions is that foreign debt accounts for only a small portion of the total debt in our sample. The mean (median) ratio of foreign debt to total debt is only
29
Loans from government-controlled banks versus loans from independent banks: La Porta, Lopez-de
Silanes, and Shleifer (2002) provide some evidence that lending decisions by government-controlled
banks in emerging markets are not value-driven, suggesting that debt from government-controlled banks
is a less effective monitoring device than debt from independent banks. To examine whether the influence
of loans in restructuring decisions is different, depending upon whether banks are government-owned or
independent, we divide bank loans into loans from government-controlled banks and those from other
banks. Because the Korean government had full ownership of the Korea Development Bank and more
than 60% of the Industrial Bank of Korea during our sample period, for our analysis we consider these two
banks to be government-controlled banks. The results show that only the ratio of loans from independent
banks to total assets is significantly and positively related to the likelihood of restructuring, supporting the
results of La Porta, Lopez-de Silanes, and Shleifer (2002).
Export ratio: During a financial crisis, firms with a high proportion of foreign sales might have more
diversified investment opportunities in foreign countries and, thus, their investment opportunities are less
likely to be adversely affected by the economic shock. Consequently, these firms might have fewer
incentives to restructure compared to firms that depend mostly on domestic sales. To address this issue,
we divide the sample into high- and low-export firms, according to the sample median, and then
re-estimate the logit regressions in Table 3 separately for these two subsamples. We find that the
significance of our key variables is similar between high- and low-export firms. These results suggest that
the effects of ownership discrepancy and leverage on restructuring are similar whether firms have large
foreign sales or not. However, liquidity-related variables are significant only among low-export firms,
suggesting that these firms were more financially-constrained during the crisis than high-export firms
were.
• Determinants of expansionary actions
To further examine the importance of owner-manager incentives and firm-specific measures of
7.93% (3.67%).
30
corporate governance in firms’ restructuring decisions in response to the economic shock, we search
expansionary actions our sample firms undertake and estimate multivariate logit regressions in which the
dependent variable is set to one if a particular expansionary action occurs, and zero otherwise.
Expansionary actions include joint ventures, strategic alliances, acquisitions, construction of new facilities,
expansion of existing production facilities, establishment of new subsidiaries or new businesses, increased
R&D or output, and financial investments in unaffiliated and affiliated firm. Out of 580 sample firms
experiencing the economic shock, we find that 234 (40%) engage in expansionary actions. The likelihood
of taking expansionary actions is positively related to the separation of the cash flow rights and control
rights of chaebol owner-managers. In contrast, the likelihood of taking expansionary actions is negatively
related to leverage and bank loan ratios. These results further suggest that owner-manager incentives and
firm-specific measures of corporate governance play an important role in making firms’ expansionary
action decisions during an economic shock.
• Effect of restructuring actions on the long-term operating performance
Previous studies show that restructuring activities caused by performance shocks are associated with
an improvement in operating performance (John and Ofek, 1995; Kang and Shivdasani, 1997; Denis and
Kruse, 2000). To address this issue, we examine the effect of restructuring actions on the long-term
operating performance of our sample firms. Specifically, we estimate the industry-adjusted ratio of
operating income to total assets (ROA) for the three years following the announcements. For the sample
of restructuring actions during the economic shock, the average change in ROA is a marginally significant
2.00% over the three-year period. When we divide the total number of sample firms into chaebol and
non-chaebol groups, the significance disappears for both chaebol firms and non-chaebol firms. For the
sample of restructuring actions during a performance decline, however, the average change in ROA is
significantly positive only for non-chaebol firms. Thus, restructuring actions result in improvements in
operating performance, especially those taken by non-chaebol firms following performance declines.
These results are consistent with those of Kang and Shivdasani (1997) and Denis and Kruse (2000). The
fact that restructuring actions during the economic shock do not improve long-term operating performance
31
for both chaebol and non-chaebol firms suggests that these actions mainly are geared toward avoiding
short-term liquidity constraints, rather than improving long-term operational efficiencies.
5. Summary and conclusions
In this paper, we examine how owner-manager incentives and firm-specific measures of corporate
governance affect restructuring decisions during an economy-wide shock. Since business environments
and claimholder incentives differ substantially during economic and performance shock periods, firms
experiencing an economic shock are likely to have different restructuring incentives compared to those
experiencing a performance shock. Consistent with this view, we find that Korean chaebol firms in the
economic shock sample display an abnormally-high frequency of restructuring actions compared to those
in the performance decline sample. Moreover, the frequency of cash-generating restructuring actions is
higher for firms in the economic shock sample than for firms in the performance decline sample,
suggesting that firms experiencing an economy shock have stronger incentives to choose restructuring that
mitigates their liquidity constraints than firms experiencing performance declines do.
We also find that, during an economic shock, the likelihood of restructuring increases with leverage,
especially bank loans, and equity ownership by unaffiliated financial institutions. Chaebol affiliation also
increases the likelihood of restructuring, but mostly for firms with high leverage. In contrast, chaebol
firms with larger ownership discrepancy are less likely to restructure than other firms are. For a pooled
sample of firms in the economic shock and performance decline samples, we find that these results are
evident mainly among firms experiencing an economic shock.
Finally, we find that, during an economic shock, chaebol firms that engage in (cash-generating)
restructuring actions realize a positive and significant announcement return, whereas during a
performance decline, both chaebol firms and non-chaebol firms experience insignificant positive returns.
The cross-sectional regression analysis shows that announcements of restructuring actions during an
economic shock by chaebol firms with a divergence in cash flow rights and control rights of controlling
shareholders and those by chaebol firms with a lower leverage ratio are greeted more negatively by
investors. Consistent with the market’s ex-ante valuation of intra-group propping, we also find that
32
abnormal returns of the chaebol restructuring firms are positively related to abnormal returns of the
equally-weighted portfolio of other non-restructuring affiliates in the same group. The sensitivity of
abnormal returns for the equally-weighted portfolio of non-restructuring affiliates to abnormal returns for
the restructuring firms is lower if the cash flow rights of the controlling shareholders in the restructuring
firms are higher. In contrast, the sensitivity is higher if the discrepancy in cash flow rights and voting
rights of the controlling shareholders in the restructuring firms is larger.
Overall, these results are consistent with the view that the separation of cash flow rights and control
rights of controlling shareholders increases their incentive to expropriate other investors, particularly
during an economic shock. However, firm-specific measures of corporate governance such as total debt,
bank loans, and equity ownership by unaffiliated financial institutions mitigate such expropriation
incentives, thereby influencing firms to choose value-maximizing restructuring policies. Our findings
that corporate governance plays a more important role in determining a firm’s restructuring decisions
during an economic shock than during a performance decline are also consistent with the view that the role
of corporate governance in firms’ investment decisions becomes more critical when interests of various
claimholders diverge more. To the extent that the divergence of interests of claimholders is likely to be
particularly severe during an economic shock, because of financial difficulties experienced by almost all
claimholders, our findings based on Korean experience are likely to have important and relevant
implications for the role of corporate governance during other economy-wide shocks, such as the global
financial crisis of 2007-2008.
Acknowledgements
We are grateful for comments from Wei-Lin Liu and the seminar participants at the 2008 American
Economic Association Meeting, the 2009 Financial Management Association Meeting, Concordia
University, Michigan State University, the National University of Singapore, and San Jose State
University. Most work for this paper has been done while Lee was at the National University of
Singapore. This paper contains the opinions of the authors, not those of Dimensional or its affiliates.
33
References Allayannis, G., Brown, G. W., Klapper, L. F., 2003. Capital structure and financial risk: Evidence from
foreign debt use in East Asia. Journal of Finance 58, 2667-2709. Bae, K.-H., Kang, J.-K., Kim, J.M., 2002. Tunneling or value added? Evidence from merger by Korean
business groups. Journal of Finance 57, 2695-2740. Baek, J.-S., Kang, J.-K., Lee, I., 2006. Tunneling and business groups: Evidence from private securities
offerings by Korean chaebols, Journal of Finance 61 2415-2449. Baek, J.-S., Kang, J.-K., Park, K.S., 2004. Corporate governance and firm value: Evidence from Korean
financial crisis. Journal of Financial Economics 71, 265-313. Claessens, S., Djankov, S., Fan, J.P.H, Lang, L.H.P., 2002. Disentangling the incentive and entrenchment
effects of large shareholders. Journal of Finance 57, 2741-2771. Claessens, S., Djankov, S., Lang, L.H.P., 2000. The separation of ownership and control in East Asian
Corporations. Journal of Financial Economics 58, 81-112. Demsetz, H., Lehn, K.M., 1985. The structure of corporate ownership: Cause and consequence. Journal of
Political Economy 93, 375-390. Denis, D., Kruse, T., 2000. Managerial discipline and corporate restructuring following performance
decline. Journal of Financial Economics 55, 391-424. Denis, D., McConnell, J., 2003. International corporate governance. Journal of Financial and Quantitative
Analysis 38, 1-36. Erkens, D., Hung, M., Matos, P., 2009. Corporate governance in the 2007-2008 financial crisis: Evidence
from financial institutions worldwide. Working paper, University of Southern California. Fama, E., 1985, What’s different about banks? Journal of Monetary Economics 10, 10-19. Gilson, S., John, K., Lang, L.H.P., 1990. Troubled debt restructurings: An empirical study of private
reorganization of firms in default. Journal of Financial Economics 27, 315-354. Harvey, C.R., Lins, K.V., Roper, A.H., 2004. The effect of capital structure when expected agency costs
are extreme. Journal of Financial Economics 74, 3-30. Jain, P.C., 1985. The effect of voluntary sell-off announcements on shareholders wealth. Journal of
Finance 40, 209-224. Jensen, M.C., 1986. Agency costs of free cash flows, corporate finance, and takeovers. American
Economic Review 76, 323-329. Jensen, M.C., 1989. Eclipse of the public corporation. Harvard Business Review 5, 61-74. Joh, S.W., 2003. Corporate governance and firm profitability: Evidence from Korea before the economic
crisis. Journal of Financial Economics 68, 287-322.
34
John, K., Lang, L.H.P., Netter, J., 1992. The voluntary restructuring of large firms in response to
performance decline. Journal of Finance 47, 891-917. John, K., Ofek, E., 1995. Asset sales and increases in focus. Journal of Financial Economics 37, 105-126. Johnson, S., Boone, P., Breach, A., Friedman, E., 2000. Corporate governance in the Asian financial
crisis. Journal of Financial Economics 58, 141-186. Kang, J.-K., 1998. Bank-centered corporate governance systems. Economic Analysis 4 (in Korean), Bank
of Korea, 120-161. Kang, J.-K., Lee, I., Na, H. S., 2004. Corporate restructuring during performance declines: The role of
ownership and corporate governance in restructuring activities and post-restructuring performance. Korean Journal of Finance 17 (in Korean), 1-40.
Kang, J.-K., Shivdasani, A., 1997. Corporate restructuring during performance decline in Japan. Journal
of Financial Economics 46, 29-65. La Porta R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of
Finance 54, 471-517. La Porta R., Lopez-de-Silanes, F., Shleifer, A., 2002. Government ownership of banks. Journal of Finance
57, 265-301. Lemmon, M.L., Lins, K., 2003. Ownership structure, corporate governance, and firm value: Evidence
from the East Asian financial crisis. Journal of Finance 58, 1445-1468. Mitton, T., 2002. A cross-firm analysis of the impact of corporate governance on the East Asian financial
crisis. Journal of Financial Economics 64, 215-241. Ofek, E., 1993. Capital structure and firm response to poor performance: An empirical analysis. Journal of
Financial Economics 34, 3-30. Powers, E. A., 2005. Interpreting logit regressions with interaction terms: An application to the
management turnover literature. Journal of Corporate Finance 11, 504-522. Shin, H.H., Park, Y.S., 1999. Financing constraints and internal capital markets: Evidence from Korean
chaebols. Journal of Corporate Finance 5, 169-191. Shleifer, A., Vishny, R., 1986. Large shareholders and corporate control. Journal of Political Economy 94,
461-488.
35
Table 1 Descriptive statistics of firm characteristics
Economic Shock Sample Performance Decline Sample
Total firms N= 580
Chaebol firms
N= 131
Non-chaebol firms
N= 449
Test of Difference Total firms N= 193
Chaebol firms N= 33
Non-chaebol firms
N= 160
Test of Difference
t-test Wilcoxon z-test t-test Wilcoxon
z-test Cash-flow rights of controlling shareholders (%)
24.42*** [24.34]***
11.92*** [8.15]***
27.94** [27.54]** 0.00 0.00 28.55
[27.33] 18.07
[19.59] 30.62
[28.87] 0.00 0.00
Voting rights of controlling shareholders (%)
29.55** [28.58]***
25.34** [24.00]**
30.74 [29.71] 0.00 0.00 32.43
[31.70] 31.36
[29.16] 32.65
[32.08] 0.61 0.65
Voting rights / Cash-flow rights 3.397*** [1.000]**
10.672*** [2.827]*
1.349** [1.000] 0.00 0.00 1.483
[1.000] 3.202
[1.498] 1.144
[1.000] 0.01 0.00
Equity ownership by unaffiliated financial institutions (%)
8.45 [6.20]
14.12 [13.00]
6.80 [4.94] 0.00 0.00 8.12
[6.13] 13.21
[12.60] 7.11
[5.30] 0.00 0.00
Total debt / total assets 0.691*** [0.689]**
0.764 [0.770]
0.670*** [0.653]** 0.00 0.00 0.642
[0.648] 0.776
[0.786] 0.615
[0.609] 0.00 0.00
Bank loans / total assets 0.312*** [0.277]***
0.308 [0.278]*
0.313*** [0.277]*** 0.76 0.43 0.234
[0.209] 0.255
[0.198] 0.230
[0.209] 0.42 0.60
Market value of equity (billion won)
130.45 [51.84]***
333.26 [142.50]
71.66* [41.30]*** 0.00 0.00 143.08
[37.48] 604.57
[125.44] 53.85
[32.11] 0.20 0.00
Tobin’s q 1.088 [1.004]
1.007** [0.953]***
1.111 [1.035] 0.00 0.00 1.097
[1.070] 1.106
[1.071] 1.095
[1.069] 0.81 0.83
Cash flow (operating income + depreciation) / total assets
0.056 [0.056]
0.057 [0.058]
0.055 [0.055] 0.71 0.63 0.057
[0.052] 0.071
[0.052] 0.054
[0.053] 0.31 0.85
Liquid assets (cash + marketable securities) / total assets
0.096** [0.074]*
0.052 [0.033]**
0.109* [0.090] 0.00 0.00 0.114
[0.084] 0.054
[0.041] 0.126
[0.091] 0.00 0.00
Export / sales 0.253*** [0.191]*
0.276 [0.239]
0.247*** [0.176]** 0.24 0.16 0.322
[0.226] 0.294
[0.260] 0.327
[0.218] 0.58 0.68
The economic shock sample includes 580 nonfinancial firms listed on the Korean Stock Exchange (KSE) as of November 18, 1997, which is the date when the IMF proposed a rescue package to Korea to help overcome the financial crisis that had started in the middle of that month. The performance decline sample includes 193 nonfinancial firms listed on the KSE that experience a more than 50% drop in their operating income (ROA) within a year during the 1994-1996 period. Firms that are in a supervision of the bankruptcy court before a shock (performance decline) are excluded from the sample. Chaebol firms are those belonging to one of the 30 largest business groups in Korea. The cash flow right of the controlling shareholder is measured as the sum of the direct equity ownership of the controlling shareholders and their family and the product of the ownership stakes obtained indirectly along the chain in a pyramid structure (by tracing up to two layers of control chains). Voting rights are taken as the sum of direct equity ownership of the controlling shareholders and their family and equity ownership by affiliated firms. All accounting variables are measured at the end of the fiscal year immediately before November 18, 1997 (performance decline year). Tobin’s q is calculated by dividing the sum of market value of equity and book value of debt by total assets. The median is reported in square brackets. The numbers in the test-of-difference columns denote p-values from the test of difference between chaebol and non-chaebol firms. The difference in each variable between the economic shock sample and the performance decline sample is tested using the t-test and the Wilcoxon z-test, and the symbols *, **, and *** in the economic shock sample columns denote significance at the 10%, 5%, and 1% levels, respectively.
36
Table 2 Frequency of restructuring actions during the economic shock and the performance decline
Economic Shock Sample Performance Decline Sample
Total firms N= 580
Chaebol firms (A) N= 131
Non-chaebol firms (B) N= 449
Test of Difference
(A-B)
Total firms N= 193
Chaebol firms (C) (N= 33)
Non-chaebol firms (D) (N= 160)
Test of Difference
(C-D)
Number (%) Number (%) Number (%) p-value of Chi-square test Number (%) Number (%) Number (%) p-value of
Chi-square test Total restructuring actions (1 + 2) 195 (33.6)* 88 (67.2)*** 107 (23.8) 0.00 51 (26.4) 14 (42.4) 37 (23.1) 0.02 1. Asset contraction actions 178 (30.7) 79 (60.3)** 99 (22.1) 0.00 48 (24.9) 13 (39.4) 35 (21.9) 0.03 By types: (1) Asset sales 151 (26.0) 69 (52.7)** 82 (18.3) 0.00 41 (21.2) 11 (33.3) 30 (18.8) 0.06 Asset sales to affiliated firms 37 (6.4)* 23 (17.6) 14 (3.1) 0.00 6 (3.1) 4 (12.1) 2 (1.3) 0.00 Asset sales to non-affiliated firms 126 (21.7) 53 (40.5)** 73 (16.3) 0.00 36 (18.7) 7 (21.2) 29 (18.1) 0.68 (2) Investment cut 14 (2.4) 6 (4.6) 8 (1.8) 0.07 2 (1.0) 1 (0.6) 1 (0.6) 0.21 (3) Closing plant or suspending operations 18 (3.1) 7 (5.3) 11 (2.5) 0.09 4 (2.1) 0 (0.0) 4 (2.5) 0.35 (4) Withdrawal from line of business 12 (2.1) 7 (5.3) 5 (1.1) 0.00 3 (1.6) 2 (6.1) 1 (0.6) 0.02 (5) Spin off 22 (3.8)** 19 (14.5)** 3 (0.7) 0.00 1 (0.5) 0 (0.0) 1 (0.6) 0.65 By cash generation: (1) Asset downsizing with cash generation 156 (26.9) 72 (55.0)*** 84 (18.7) 0.00 41 (21.2) 10 (30.3) 31 (19.4) 0.16 (2) Asset downsizing without cash generation 48 (8.3) 25 (19.1) 23 (5.1) 0.00 10 (5.2) 4 (12.1) 6 (3.8) 0.05 2. Employment changes 44 (7.6)** 28 (21.4)** 16 (3.6) 0.00 5 (2.6) 2 (6.1) 3 (1.9) 0.17 (1) Layoffs 28 (4.8)** 16 (12.2) 12 (2.7) 0.00 3 (1.6) 1 (3.0) 2 (1.3) 0.45 (2) Pay cut 19 (3.3)** 12 (9.2)* 7 (1.6) 0.00 1 (0.5) 0 (0.0) 1 (0.6) 0.65 (3) Lowering mandatory retirement ages or
offering early retirement incentives 3 (0.5) 3 (2.3) 0 (0.0) 0.00 1 (0.5) 1 (3.0) 0 (0.0) 0.03
The economic shock sample includes 580 nonfinancial firms listed on the Korean Stock Exchange (KSE) as of November 18, 1997, which is the date when the IMF proposed a rescue package to Korea to help overcome the financial crisis that had started in the middle of that month. The performance decline sample includes 193 nonfinancial firms listed on the KSE that experience a more than 50% drop in their operating income (ROA) within a year during the 1994-1996 period. Firms that are in a supervision of the bankruptcy court before a shock (performance decline) are excluded from the sample. Chaebol firms are those belonging to one of the 30 largest business groups in Korea. To identify restructuring activities of firms, we search the Korea Investor’s Network for Disclosure system and the Korean Economic Daily Newspaper. For the economic shock sample, all sources are examined for stories during the one-year period starting from November 18, 1997. For the performance decline sample, we examine the sources for an one-year period from six months before to six months after the fiscal year end of the poor-performance year. Number indicates the number of firms that announce each action and % in parentheses indicates the percentage of firms that announce each action out of total firms in each category. The numbers in the test-of-difference columns denote p-values for the Chi-square test of the difference in mean percentage of firms between chaebol and non-chaebol firms. The difference in mean percentage between the economic shock sample and the performance decline sample is tested using the Chi-square test, and the symbols *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.
37
Table 3 Logit estimates of the likelihood of restructuring actions during the economic shock (1) (2) (3) (4) (5)
Chaebol dummy: (A) 0.8817** (0.01)
0.4502 (0.22)
0.4852 (0.20)
Log (voting rights / cash flow rights): (B) 0.4765* (0.07)
0.1315 (0.47)
0.1387 (0.45)
(A) × (B) -0.6705* (0.06)
Total debt / total assets: (C) 2.2357*** (0.00)
1.9509*** (0.00) 2.0953***
(0.00)
(A) × (C) 4.5397** (0.05)
Bank loans / total assets: (D) 2.3217*** (0.00) 2.3159***
(0.00)
Non-bank debt / total assets: (E) 1.3902* (0.10) 1.6443**
(0.04)
(A) × (D) 5.6551** (0.03)
(A) × (E) 4.3335* (0.08)
Dummy for chaebol firms with ownership discrepancy (equals one if the ratio of voting rights to cash flow rights of the chaebol firm > 1): (F)
0.1862 (0.57)
0.2834 (0.42)
(F) × (G) 3.0938 (0.21)
(F) × (H) 4.7538* (0.10)
(F) × (I) 2.3972 (0.36)
Equity ownership by unaffiliated financial institutions 2.7655* (0.07)
2.8803* (0.06)
3.2235** (0.04)
2.9440** (0.04)
3.2817** (0.03)
Log (market value of equity) 0.5596*** (0.00)
0.5537*** (0.00)
0.5593*** (0.00)
0.6529*** (0.00)
0.6530*** (0.00)
Tobin’s q -1.3866*** (0.00)
-1.2932*** (0.01)
-1.3177*** (0.01)
-1.4940*** (0.00)
-1.5060*** (0.00)
Cash flow (operating income + depreciation) / total assets 0.4661 (0.85)
0.5420 (0.83)
0.3460 (0.89)
0.3239 (0.90)
0.3323 (0.89)
Liquid assets (cash + marketable securities) / total assets -1.6390 (0.31)
-1.7361 (0.28)
-1.9929 (0.22)
-2.5960* (0.10)
-2.7820* (0.08)
Export / Sales -0.2905 (0.53)
-0.2758 (0.55)
-0.2573 (0.57)
-0.2343 (0.60)
-0.1962 (0.66)
Intercept -10.9850*** (0.00)
-9.5004*** (0.00)
-9.5826*** (0.00)
-10.9003*** (0.00)
-10.9279*** (0.00)
Industry dummies YES YES YES YES YES
Number of firms involved in restructuring actions 186 186 186 188 187
Number of firms in sample 556 556 555 563 561
Model p-value 0.00 0.00 0.00 0.00 0.00 The sample includes 580 nonfinancial firms listed on the Korean Stock Exchange (KSE) as of November 18, 1997, which is the date when the IMF proposed a rescue package to Korea to help overcome the financial crisis that had started in the middle of that month. Firms that are in a supervision of the bankruptcy court before a shock are excluded from the sample. To identify restructuring activities of firms, we search the Korea Investor’s Network for Disclosure system and the Korean Economic Daily Newspaper. The sources are examined for stories during the one-year period starting from November 18, 1997. The dependent variable equals one if firms undertake at least one of the restructuring actions categorized in Table 2. The chaebol dummy takes the value of one if firms are those belonging to one of the 30 largest business groups in Korea. The cash flow right of the controlling shareholder is measured as the sum of the direct equity ownership of the controlling shareholders and their family and the product of the ownership stakes obtained indirectly along the chain in a pyramid structure (by tracing up to two layers of control chains). Voting rights are taken as the sum of direct equity ownership of the controlling shareholders and their family and equity ownership by affiliated firms. Tobin’s q is calculated by dividing the sum of market value of equity and book value of debt by total assets. All accounting variables are measured at the end of the fiscal year immediately before November 18, 1997. All variables that are interacted with a dummy variable are demeaned. Numbers in parentheses are p-values. The symbols *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.
38
Table 4 Cumulative abnormal returns [CAR (-1, 1)] around the announcement of restructuring activities
Panel A: CARs (-1, 1) for the sample of restructuring actions during the economic shock
Total sample (A + B)
Restructuring with cash generation (A)
Restructuring without cash generation (B)
Test of difference (A – B)
Total sample (1 + 2)
0.0133 (1.50)
[0.016]* (n=149)
0.0231** (2.26)
[0.023]*** (n=121)
-0.0287* (-1.91)
[-0.023] (n=28)
0.052*** (2.85)
[0.047]***
Chaebol firms (1)
0.0342*** (2.75)
[0.0252]*** (n=66)
0.0457*** (3.25)
[0.0262]*** (n=54)
-0.0177 (-0.82)
[-0.0223] (n=12)
0.063** (2.01)
[0.049]*
Non-chaebol firms (2)
-0.0032 (-0.26) [0.005] (n=83)
0.0049 (0.34)
[0.016] (n=67)
-0.0371* (-1.75)
[-0.024] (n=16)
0.042 (1.35)
[0.040]*
Test of difference (1 – 2) 0.037** (2.11)
[0.020]*
0.041** (2.01)
[0.010]
0.019 (0.63)
[0.002]
Panel B: CARs (-1, 1) for the sample of restructuring actions during a performance decline
Total sample (A + B)
Restructuring with cash generation (A)
Restructuring without cash generation (B)
Test of difference (A – B)
Total sample (1 + 2)
0.0165* (1.77)
[0.0052]* (n=40)
0.0196* (1.73)
[0.0061]* (n=32)
0.0040 (0.37)
[0.0021] (n=8)
0.016 (1.00)
[0.004]
Chaebol firms (1)
0.0270 (1.03)
[0.0054] (n=11)
0.0439 (1.27)
[0.0129] (n=8)
-0.0183 (-1.76)
[-0.0286] (n=3)
0.062 (1.06)
[0.042]
Non-chaebol firms (2)
0.0125 (1.47)
[0.0052] (n=29)
0.0115 (1.15)
[0.0054] (n=24)
0.0174 (1.33)
[0.0052] (n=5)
-0.006 (0.26)
[0.000]
Test of difference (1 - 2) 0.014 (0.69)
[0.000]
0.032 (1.25)
[0.008]
-0.036 (1.87)
[-0.0338]
The economic shock sample includes 580 nonfinancial firms listed on the Korean Stock Exchange (KSE) as of November 18, 1997, which is the date when the IMF proposed a rescue package to Korea to help overcome the financial crisis that had started in the middle of that month. The performance decline sample includes 193 nonfinancial firms listed on the KSE that had experienced more than a 50% drop in their operating income (ROA) within a year during the 1994-1996 period. Firms that are in a supervision of the bankruptcy court before the shock (performance decline) are excluded from the sample. We identify initial public announcement dates of the restructuring activities from the Korea Investor’s Network for Disclosure system, and a daily newspaper, the Korean Economic Daily Newspaper. We use as the announcement date the date that a news announcement first appeared in either of these two publications. We exclude observations that accompany other restructuring announcements during the estimation period of the market model. For each event, we compute the abnormal return using the market model. We use the KOSPI return as the market return proxy. We obtain our estimates of the market model by using 200 trading days of return data beginning 220 days and ending 21 days before the restructuring announcement. We then calculate daily abnormal returns and sum the daily abnormal returns over three days around the restructuring announcement date to get the cumulative abnormal return (CAR). Chaebol firms are those belonging to one of the 30 largest business groups in Korea. The mean CAR (-1, 1) is reported on top, followed by t-statistic in parentheses, median CAR (-1, 1) in square bracket, and the number of observations in parentheses at the bottom. The symbols *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.
39
Table 5 OLS regression of cumulative abnormal returns [CAR (-1, 1)] around the announcement of cash-generating restructuring activities during the economic shock on explanatory variables (1) (2) (3) (4)
Chaebol dummy: (A) -0.0061 0.0231 (0.89) (0.46)
Log (voting rights / cash flow rights): (B) 0.1240* -0.0110 (0.08) (0.60)
(A) × (B) -0.1380* (0.06)
Total debt / total assets: (C) -0.0473 -0.0897 -0.0404 (0.54) (0.28) (0.60)
(A) × (C) 0.3893** (0.04)
Dummy for chaebol firms with ownership discrepancy (equals one if the ratio of voting rights to cash flow rights of the chaebol firm > 1): (D)
-0.0222 -0.0243
(0.44) (0.41)
(C) × (D) 0.4082* (0.08)
Bank loans / total assets: (E) -0.0480 (0.56)
Non-bank debt / total assets: (F) -0.0203 (0.83)
(D) × (E) 0.4020* (0.10)
(D) × (F) 0.4619* (0.09)
Equity Ownership by unaffiliated financial institutions 0.0149 0.0485 0.0798 0.0719 (0.91) (0.73) (0.55) (0.60)
Log (market value of equity) 0.0064 0.0103 0.0130 0.0127 (0.55) (0.33) (0.19) (0.20)
Tobin’s q -0.0052 0.0008 -0.0003 0.0048 (0.86) (0.98) (0.99) (0.87)
Cash flow (operating income + depreciation) / total assets 0.2819 0.3768 0.3783 0.3752 (0.23) (0.11) (0.11) (0.12)
Liquid assets (cash + marketable securities) / total assets -0.0245 0.0120 -0.0246 -0.0040 (0.89) (0.95) (0.89) (0.98)
Intercept -0.0305 -0.2099 -0.2422 -0.2427 (0.89) (0.26) (0.16) (0.17)
Industry dummies YES YES YES YES Number of observations 119 119 120 120 Adjusted R2 0.048 0.053 0.034 0.020
The sample includes 580 nonfinancial firms listed on the Korean Stock Exchange (KSE) as of November 18, 1997, which is the date when the IMF proposed a rescue package to Korea to help overcome the financial crisis that had started in the middle of that month. Firms that are in a supervision of the bankruptcy court before a shock are excluded from the sample. To identify restructuring activities of firms, we search the Korea Investor’s Network for Disclosure system and the Korean Economic Daily Newspaper. The sources are examined for stories during the one-year period starting from November 18, 1997. The dependent variable is the cumulative abnormal return (CAR) calculated in Table 5. The chaebol dummy takes the value of one if firms are those belonging to one of the 30 largest business groups in Korea. The cash flow right of the controlling shareholder is measured as the sum of the direct equity ownership of the controlling shareholders and their family and the product of the ownership stakes obtained indirectly along the chain in a pyramid structure (by tracing up to two layers of control chains). Voting rights are taken as the sum of direct equity ownership of the controlling shareholders and their family and equity ownership by affiliated firms. Tobin’s q is calculated by dividing the sum of market value of equity and book value of debt by total assets. All accounting variables are measured at the end of the fiscal year immediately before November 18, 1997. All variables that are interacted with a dummy variable are demeaned. Numbers in parentheses are p-values. The symbols *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.
40
Table 6 OLS regression of equally-weighted portfolio cumulative abnormal returns [CAR(-1,1)] of other chaebol firms that belong to the same chaebol as the restructuring firm Panel A: Equally-weighted portfolio CAR(-1, 1) of non-restructuring firms that belong to the same chaebol as the restructuring firm
Total sample
Restructuring with cash generation (A)
Restructuring without cash generation (B)
Test of difference (A – B)
0.0240*** (2.61)
[0.0116]** (n=65)
0.0318*** (3.01)
[0.0174]*** (n=53)
-0.0108 (-0.79)
[-0.0145] (n=12)
0.043* (1.83)
[0.032]*
Panel B: OLS regression of equally-weighted portfolio CAR(-1, 1) of non-restructuring firms that belong to the same chaebol as the restructuring firm on restructuring firms’ financial characteristics (1) (2) (3) (4)
Intercept 0.1641 (0.30)
0.1532 (0.35)
-0.0193 (0.90)
0.0684 (0.67)
CAR (-1,1) for restructuring firms: (A) 0.5290*** 0.6696*** 0.5300*** 0.4385*** (0.00) (0.00) (0.00) (0.00)
Cash flow rights of controlling shareholders -0.2753*** (0.00)
Dummy for high cash flow rights (equals one if cash flow rights of the controlling shareholders are greater than the sample median): (B)
-0.0464*** (0.00)
(A) × (B) -0.2838* (0.07)
Log (voting rights / cash flow rights) 0.0124** (0.05)
Dummy for high discrepancy ratio (equals one if the ratio of voting rights to cash flow rights is greater than the sample median): (C)
0.0416*** (0.01)
(A) × (C) 0.2575* (0.09)
Total debt / total assets 0.1217 0.0859 0.1066 0.0759 (0.16) (0.34) (0.27) (0.41)
Equity Ownership by unaffiliated financial institutions 0.0010 0.0638 0.0666 0.0840 (0.99) (0.42) (0.43) (0.30)
Dummy for cash-generating restructuring actions 0.0085 0.0034 0.0076 0.0040 (0.65) (0.87) (0.71) (0.84)
Log (market value of equity) -0.0128** -0.0096 -0.0051 -0.0071 (0.05) (0.14) (0.44) (0.27)
Tobin’s q -0.0224 -0.0255* -0.0267* -0.0268* (0.12) (0.09) (0.10) (0.08)
Cash flow (operating income + depreciation) / total assets 0.4127* 0.3000 0.4200* 0.2828 (0.06) (0.19) (0.08) (0.23)
Liquid assets (cash + marketable securities) / total assets 0.1526 0.0884 0.0814 0.0633 (0.30) (0.56) (0.62) (0.68)
Industry dummies YES YES YES YES Number of observations 64 64 64 64 Adjusted R2 0.538 0.509 0.439 0.492
The sample comprises 65 chaebol firms that engage in restructuring actions between November 18, 1997 and November 17, 1998 and 65 portfolios of other chaebol firms that do not engage in restructuring actions during the same period. Abnormal returns for other group affiliates are calculated by using the market model. The market model parameters are estimated using returns of the equally-weighted portfolio of other firms that belong to the same chaebol as the restructuring firm and then used to calculate the daily abnormal returns of the portfolio. The cash flow right of the controlling shareholder is measured as the sum of the direct equity ownership of the controlling shareholders and their family and the product of the ownership stakes obtained indirectly along the chain in a pyramid structure (by tracing up to two layers of control chains). Voting rights are taken as the sum of direct equity ownership of the controlling shareholders and their family and equity ownership by affiliated firms. Tobin’s q is calculated by dividing the sum of market value of equity and book value of debt by total assets. All accounting variables are measured at the end of the fiscal year immediately before November 18, 1997. In Panel A, the mean equally-weighted portfolio CAR (-1, 1) is reported on top, followed by t-statistic in parentheses, median equally-weighted portfolio CAR (-1, 1) in square bracket, and the number of observations in parentheses at the bottom. In Panel B, numbers in parentheses are p-values. The symbols *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.